In the past, when semiconductor and power-supply companies discussed the motivations for putting digital control in the loop, power management requirements were usually cited as the main justification. Even companies that develop digital power controllers will sometimes argue that it doesn't matter to the customer whether the control loop is implemented digitally. However, that approach opens the door to implementing all types of functions that would otherwise require many external components.
That reduction in component count is particularly helpful when implementing complex, distributed power architectures. Moreover, having a digital power controller that is easily configurable for output-voltage levels, tracking and other parameters makes it much easier to make changes when designing the power system.
Although there's no doubt that such benefits are compelling reasons for digital control, perhaps the industry has focused on those benefits because they represent the most pressing needs. Nevertheless, many feel that the focus on power-management functions puts digital technology into a position where it's simply emulating analog power control.
But as the industry continues its development of digital power components and solutions, more attention is being given to the use of digital control techniques to optimize power conversion and perform functions that typically aren't feasible with analog techniques. Ultimately, such functions may represent the true value of digital, as they give power system designers new options for improving fundamental metrics of power-supply performance such as efficiency and transient response.
This supplement to Power Electronics Technology sheds some light on the design techniques that can be used to implement such functions. For example, in “Digital Power Control Enables System Identification,” authors Brett Etter and Ross Fosler of Silicon Labs describe how the implementation of digital control provides the designer with “complete visibility into the entire control loop to measure frequency response in real time without the use of a network analyzer.”
This capability becomes a development tool that power-supply designers can exploit when fine tuning their designs on the bench. However, the same capability can be extended to allow for optimization of power-supply performance in the application. That concept is explored in “Digital Control Measures In-System Response.” In this article, Mark Hagen and Dave Freeman of Texas Instruments describe a technique for measuring the transfer function of the power stage with an eye toward building systems that can perform auto tuning, diagnostics and on-the-fly adaptability.
In each of these articles, the actual implementation of the design is built around a digital signal processor (DSP) or controller (DSC). In another feature, “DSCs Ease Migration to Digital Loop Control,” Bryan Kris of Microchip Technology discusses the broader issues surrounding the use of DSCs in power-supply designs. Here, Kris makes the case that “recent advances in digital signal controllers with on-chip high-performance peripherals targeting power conversion, coupled with their ease of use and affordability enable many more power conversion products to migrate to digital loop control.” In this feature, Kris discusses the fundamental design issues that arise when using DSCs to implement digital loop control.
Although DSPs and DSCs promise the greatest flexibility for designers looking to build advanced features into their supplies, the state-machine-based approach to digital control also offers some advanced capabilities. In “Explore The Lesser-Known Benefits of Digital Power,” Marty Pandola of Zilker Labs explains how even the more hardwired approach to digital control permits tuning of the power-conversion stage, enabling functions such as programming of output-voltage droop for improved transient response.
In a sense, the title of Pandola's article is one that could apply to all of the features in this issue, as all the authors aim to shed light on design possibilities that many engineers have yet to explore.